[donunus]

Quote:

Originally Posted by xnor /img/forum/go_quote.gif Why? HeadRoom doesn't mod them before measuring, right.

Thats not the point. I was saying that pads and enclosures are a part of every headphone and as you can see, any change in even just the pads(what more if you change the enclosures) even yields a big change.

to add to what you said...
As you can see with the sr60s response that it is more extended in the bass compared to all other grados up to rs1s. I believe this is more of a direct result of how big the pads impact the sound. Look at the huge dives of the bass slopes of the models fitted with bowls compared to the sr60s with comfies. RS1s are supposedly bassier than sr80s and even sr80s are supposedly bassier than the 60s. Then how come the 60 is the most extended? Its the comfies vs bowls were dealing with here. Im sure if the 60s were fitted with bowls, it would have the least amount of bass in the lineup.

 

[xnor]

Quote:

Originally Posted by Shike /img/forum/go_quote.gif Which isn't reliable on the FR graph.

We're going around in circles..


Quote:

Originally Posted by Shike /img/forum/go_quote.gif Control over time. Let's say you're running an FR sweep . . . it doesn't spend a lot of time on a single frequency. By using a square wave of a single frequency you can see whether the transducer is able to keep it at a consistent volume linearly. It also gives an idea of mechanical damping and whether the driver is able to recover, if it's too loose, etc.

Control over time? I don't know what you're talking about here, care to explain on the square wave graph?

Take a look at this:


I was even able to reproduce the XB700's square wave by using an EQ curve that looks nearly identical to its FR.

So my conclusion is that you can only tell how well the mechanical damping is (see the green encircled area), but you can also see that by looking at the resonances in the FR (like b0dhi said; CSD would be much better though...)
example: poorly damped grados

(FR, CSD)

Quote:

Looks like an issue with the measuring equipment. Similar results come out of a Stax rig

I was referring to the second/third order harmonics which I don't see on the graph you posted.

 

[Shike]

Quote:

Originally Posted by xnor /img/forum/go_quote.gif We're going around in circles..

Eh?

Quote:

Control over time? I don't know what you're talking about here, care to explain on the square wave graph? Take a look at this: I was even able to reproduce the XB700's square wave by using an EQ curve that looks nearly identical to its FR.

I need to see labels on what you're actually doing here to understand. Also what was your testing methodology?

Quote:

[So my conclusion is that you can only tell how well the mechanical damping is (see the green encircled area)

I suggest you read headroom's explanation, those peaks are relatively meaningless. Mechanical damping is more control over over-excursion (no sharp rise in bass).

Quote:

but you can also see that by looking at the resonances in the FR (like b0dhi said; CSD would be much better though...)

Where are you finding this resonance? After so high in terms of frequencies you have issues with cancellation among other things.

Quote:

I was referring to the second/third order harmonics which I don't see on the graph you posted.

I see, I didn't notice. Unless we know how it's measured it's largley pointless though isn't it?

 

[Bradan]

Quote:

Originally Posted by donunus /img/forum/go_quote.gif Do you know for sure if all the sound going into the flat measurement is purely from the speaker cones? Does a speaker port not make sound? Dont boxes vibrate? Does a box give good transient response on their own without the drivers? If they dont then isn't the time it takes for the sound to come out of the box vibrations delayed compared to the sound directly coming from the drivers? If you don't get my point then more power to you. Lets just add another example in the mix.... put a speaker 20 ft from you and another one 5ft away.... Use a digital eq to perfectly match the in room response of both speakers.... So now they both perfectly match each other in frequency response but have a different sound... Why may I ask is that? Its because the sound of the former has a lot more delay than the other. Most of what made it flat in response were sounds created by its environment instead of a more direct sound in the speaker 5ft away. Thats basically all I was explaining in the previous post. Simple logic. I was just explaining it in a way where it relates to everything from headphones to speakers to real live sound. and Bradan, Whats with the attitude about my post count? I never shoved my post count to you to give what I said more validity. I'm only sharing what Ive learned so take what you want and leave the rest to others that appreciate what I have to say.

Yes, I know, they tune drivers, they don't just stick them in with their natural bell-curve roll-off. This is on the frequency response graph. The box resonates little, if any, as speaker boxes are made to dampen, not reflect, and any effect will still be accounted for in the final tuning of the speaker. Headphones generally do not suffer from room contamination,and I didn't say that a speaker sounded the same out-of-phase, I said that neutral speakers sound the same as other neutral speakers in the same environment.

Neutral speakers sound neatral in a neutral environment, that's the main thing I'm saying.

Sorry if my post seemed hostile in any way, as it seemed you just self-rightiously chimed in and I just wanted to disclose that the post-count shouldn't influence the assessment of each post.

 

[xnor]

Quote:

Originally Posted by Shike /img/forum/go_quote.gif I need to see labels on what you're actually doing here to understand. Also what was your testing methodology?

You mentioned "control over time", so I was guessing that you mean how long the line can keep above 0.

About the screenshot:
What I did here was creating a perfect 50 Hz square wave, then I EQ'd it using the curve you can see on the top right. The result can be seen on the left.

The red line (high pass filter) simulates the bass roll off of headphones.
Adding a dip at around 50 Hz (blue line) will cause the square wave line to stay above 0, as can be seen on the left side.

(The red/blue lines on the EQ are a bit offside, so that the EQ curve isn't hidden by them.)

This is comparable to real-world measurements as shown in the lower part.
The only difference is in the area of the overshoot (which translates into mechanical damping).

=> If you ignore the overshoot area you're basically looking at a "different form" of the FR graph.

 

[donunus]

Quote:

Originally Posted by Bradan /img/forum/go_quote.gif Yes, I know, they tune drivers, they don't just stick them in with their natural bell-curve roll-off. This is on the frequency response graph. The box resonates little, if any, as speaker boxes are made to dampen, not reflect, and any effect will still be accounted for in the final tuning of the speaker. Headphones generally do not suffer from room contamination,and I didn't say that a speaker sounded the same out-of-phase, I said that neutral speakers sound the same as other neutral speakers in the same environment. Neutral speakers sound neatral in a neutral environment, that's the main thing I'm saying. Sorry if my post seemed hostile in any way, as it seemed you just self-rightiously chimed in and I just wanted to disclose that the post-count shouldn't influence the assessment of each post.

The key thing to remember though is that there is no such thing as a neutral environment

 

[Seamaster]

I never look at those Frequency Response graphs. I only trust my ears. wast of time

 

[Shike]

Quote:

Originally Posted by xnor /img/forum/go_quote.gif You mentioned "control over time", so I was guessing that you mean how long the line can keep above 0.

It's how straight and horizontal they can remain.

Quote:

This is comparable to real-world measurements as shown in the lower part. The only difference is in the area of the overshoot (which translates into mechanical damping). => If you ignore the overshoot area you're basically looking at a "different form" of the FR graph.

Except it doesn't immediately translate to FR. It shows you how the unit will respond to bass in general by showing how pressure builds and whether it does it linearly. I've already admitted that a CSD is more desirable, but the squarewave isn't useless by any means.

Have you read Headroom's FAQ on the 50hz squarewave?

 

[xnor]

Quote:

Originally Posted by Shike /img/forum/go_quote.gif It's how straight and horizontal they can remain.

Yeah but how long the line remains horizontal tells nothing at all about something like "control over time".

It's just a different view at the frequency response, that's why I made that screenshot.


I did look at headroom's explanation and it bascially says the same ("shows how well HPs can reproduce low frequencies").
If the FR is flat it can reproduce very low frequencies as well as higher ones, if it's rolled off it cannot. This bends the line in the square wave graph and nothing else.

And it does directly translate into FR, just not very well because of the focus on 50/500 Hz.



edit: Found another shocking distortion screeny:

That's a k271 at <90 dB. (remember that these were made to be used at high volume levels!)

Specs of course say THD <0,3 % (@1 kHz) and headroom graph doesn't look so bad either ... another reason I'd like to see THD over the full range and not just at a randomly chosen frequency.

 

[Shike]

Quote:

Originally Posted by xnor /img/forum/go_quote.gif Yeah but how long the line remains horizontal tells nothing at all about something like "control over time".

Yes it does, it shows whether the driver can produce bass notes linearly or not (preventing distortion in bass).

Quote:

It's just a different view at the frequency response, that's why I made that screenshot.

Is the upper left image an actual measurement of a transducer after the 50hz EQ'ing? If not all it shows is that a square wave can be modified, which defeats the purpose in relationship to measuring a transducer. I really don't see what you're trying to accomplish with it

I don't know how to explain this to you well. Square waves can show things like phase, break-up characteristics, and other oddities you just don't see in a frequency response.

Quote:

. . . This bends the line in the square wave graph and nothing else.

A bend in the line shows lack of linearity. Too fast of an increase or decrease in pressure tells you about the motor structure in general, and how coherent it will be to an extent.


You use a square wave as a reference to see how the transducer reacts in practice. All of thes above will impact the FR, but you can't derive these from the FR (at least not easily or in any way I know).

I suggest you read about square waves in respect to speakers here.

Quote:

edit: Found another shocking distortion screeny: That's a k271 at <90 dB. Specs of course say THD <0,3 % .. another reason I'd like to see THD over the full range and not just at a randomly chosen frequency.

Who measured it? What gear was used to measure it? What was the methodology of measuring?

Headroom shows the highest distortion product still being under 1%. I at least know how they measure.


EDIT:

Either way it looks like we're going to have to agree to disagree about the usefulness of the square wave.

 

[xnor]

Quote:

Originally Posted by Shike /img/forum/go_quote.gif Yes it does, it shows whether the driver can produce bass notes linearly or not (preventing distortion in bass).

Ok.

Quote:

Is the upper left image an actual measurement of a transducer after the 50hz EQ'ing?

No it's just an EQd wave. I didn't think of the linearity of the EQ.. but results look very close to what headroom measures.
Using a linear EQ the resulting square wave really looks a lot better.

Still, I think that there is a correlation between flat response and good looking square waves

, at least from the measurements I've seen on headroom.

Quote:

Who measured it? What gear was used to measure it? What was the methodology of measuring? Headroom shows the highest distortion product still being under 1%. I at least know how they measure.

They most certainly used quality equip (like the Earthworks M50 mic).

At what SPL does headroom make those distortion measurements?

 

[Shike]

Quote:

Originally Posted by xnor /img/forum/go_quote.gif Still, I think that there is a correlation between flat response and good looking square waves , at least from the measurements I've seen on headroom.

There is some correlation, but it's not the ultimate purpose of the square wave is what I probably should have said towards the beginning.

Quote:

They most certainly used quality equip (like the Earthworks M50 mic).

I did a translation on the page through Google. It seems they still acknowledge its relative low distortion, though it tends to crop up in bass reproduction.

Quote:

At what SPL does headroom make those distortion measurements?

90dB


Those measurements give a bit of food for thought, though. I ran Ryumatsuba's site through translation and it seems you tests 100Hz, 1K, and 10K on distortion. So it seems like his distortion measurements are more in-depth than Headroom's, but Headroom has better equipment for measuring FR itself. Going through a few decent quality dynamics on his site though I must stand by my statement: most above average headphones will have below or close to 1% THD. The K271 and a few others seem like oddballs. Compared to most dynamic loudspeakers these are still good results though